R Notebook

From lots of studies it seems that in order to capture deep meditation we need to: Do FFT (furrier transform) and check power of alfa (8-12 Hz) and theta (4-8Hz). They should increase as the meditation is deeper. especially in midline areas (electrodes in the middle of the forehead and middle of the scalp) and frontal areas (forehead). If we can check also coherence then it seems that for deeper meditation we should see higher coherence in alfa and theta in the midline and frontal areas. Maybe also we’ll see increase of activity in the right hemisphere. We need to try and check it before BM.

Coherence means to check correlation between waves. How similar are they along time. In our case they measure in the studies coherence of alpha waves from different electrodes, and coherence of theta waves from different electrodes. They found that the coherence of alpha and the coherence of theta grew as the meditation was deeper (means that alpha waves from different electrodes were more similar to each other as the meditation was deeper and same for theta waves). in the last Burning man, the most beautiful installation did just that! they measured heart bits of 6 people at once and checked their coherence and presented that with music and LED’s. As the coherence increased the music and the LED’s were increased as well. It was incredible!

---

I was working on data exploration. There are visible patterns of increasing alpha and theta channels, espacially for 1 and 2 electrodes.

Where are these two electrodes located?

I explored the data manually, and I discovered some things Nir told us about. I need more data to start with machine learning part.

print(max(eeg$timestamps) - min(eeg$timestamps))

[1] 306.8416

eeg <- read.csv('/Users/sofiagodovykh/Downloads/MuseDirectCloud_180701/Muse-B1C1_2018-06-20--07-42-37_1529495275603.csv')
library(tidyr)
library(ggplot2)

without_na = eeg[!is.na(eeg$theta_absolute_1),]
ggplot(aes(y=theta_absolute_1, na.rm=TRUE, x = timestamps), 
       data = without_na) + geom_line() + geom_smooth(color="red", span = 0.01) 

without_na = eeg[!is.na(eeg$theta_absolute_1),]
without_na = without_na[!is.na(without_na$theta_absolute_2),]
without_na = without_na[!is.na(without_na$theta_absolute_3),]
without_na = without_na[!is.na(without_na$theta_absolute_4),]
ggplot() + geom_line(aes(y=theta_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4]) + geom_smooth(aes(y=theta_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4], span = 0.01) +
 geom_line(aes(y=theta_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3]) + geom_smooth(aes(y=theta_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3], span = 0.01) +
geom_line(aes(y=theta_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2]) + geom_smooth(aes(y=theta_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2], span = 0.01) +
  geom_line(aes(y=theta_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1]) + geom_smooth(aes(y=theta_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1], span = 0.01) +
  labs(y = "theta", title = "theta")

without_na = eeg[!is.na(eeg$alpha_absolute_1),]
without_na = without_na[!is.na(without_na$alpha_absolute_2),]
without_na = without_na[!is.na(without_na$alpha_absolute_3),]
without_na = without_na[!is.na(without_na$alpha_absolute_4),]
ggplot() + geom_line(aes(y=alpha_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4]) + geom_smooth(aes(y=alpha_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4], span = 0.01) +
 geom_line(aes(y=alpha_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3]) + geom_smooth(aes(y=alpha_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3], span = 0.01) +
geom_line(aes(y=alpha_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2]) + geom_smooth(aes(y=alpha_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2], span = 0.01) +
  geom_line(aes(y=alpha_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1]) + geom_smooth(aes(y=alpha_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1], span = 0.01) +
  labs(y = "alpha", title = "alpha")

without_na = eeg[!is.na(eeg$delta_absolute_1),]
without_na = without_na[!is.na(without_na$delta_absolute_2),]
without_na = without_na[!is.na(without_na$delta_absolute_3),]
without_na = without_na[!is.na(without_na$delta_absolute_4),]
ggplot() + geom_line(aes(y=delta_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4]) + geom_smooth(aes(y=delta_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4], span = 0.01) +
 geom_line(aes(y=delta_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3]) + geom_smooth(aes(y=delta_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3], span = 0.01) +
geom_line(aes(y=delta_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2]) + geom_smooth(aes(y=delta_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2], span = 0.01) +
  geom_line(aes(y=delta_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1]) + geom_smooth(aes(y=delta_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1], span = 0.01) +
  labs(y = "delta", title = "delta")

without_na = eeg[!is.na(eeg$gamma_absolute_1),]
without_na = without_na[!is.na(without_na$gamma_absolute_2),]
without_na = without_na[!is.na(without_na$gamma_absolute_3),]
without_na = without_na[!is.na(without_na$gamma_absolute_4),]
ggplot() + geom_line(aes(y=gamma_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4]) + geom_smooth(aes(y=gamma_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4], span = 0.01) +
 geom_line(aes(y=gamma_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3]) + geom_smooth(aes(y=gamma_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3], span = 0.01) +
geom_line(aes(y=gamma_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2]) + geom_smooth(aes(y=gamma_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2], span = 0.01) +
  geom_line(aes(y=gamma_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1]) + geom_smooth(aes(y=gamma_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1], span = 0.01) +
   labs(y = "gamma", title = "gamma")

without_na = eeg[!is.na(eeg$beta_absolute_1),]
without_na = without_na[!is.na(without_na$beta_absolute_2),]
without_na = without_na[!is.na(without_na$beta_absolute_3),]
without_na = without_na[!is.na(without_na$beta_absolute_4),]
ggplot() + geom_line(aes(y=beta_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4]) + geom_smooth(aes(y=beta_absolute_1, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[4], span = 0.01) +
 geom_line(aes(y=beta_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3]) + geom_smooth(aes(y=beta_absolute_2, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[3], span = 0.01) +
geom_line(aes(y=beta_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2]) + geom_smooth(aes(y=beta_absolute_3, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[2], span = 0.01) +
  geom_line(aes(y=beta_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1]) + geom_smooth(aes(y=beta_absolute_4, x = timestamps), 
       data = without_na, color=brewer.pal(4, "Blues")[1], span = 0.01) +
   labs(y = "beta", title = "beta")

ggplot() + geom_line(aes(y=alpha_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_1),], color=brewer.pal(5, "Blues")[5], alpha = 0.5) + geom_smooth(aes(y=alpha_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_1),], color=brewer.pal(5, "Blues")[5], span = 0.01) +
geom_line(aes(y=beta_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_1),], color=brewer.pal(5, "Blues")[4], alpha = 0.5) + geom_smooth(aes(y=beta_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_1),], color=brewer.pal(5, "Blues")[4], span = 0.01) +
geom_line(aes(y=gamma_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_1),], color=brewer.pal(5, "Blues")[3], alpha = 0.5) + geom_smooth(aes(y=gamma_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_1),], color=brewer.pal(5, "Blues")[3], span = 0.01) +
geom_line(aes(y=delta_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_1),], color=brewer.pal(5, "Blues")[2], alpha = 0.5) + geom_smooth(aes(y=delta_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_1),], color=brewer.pal(5, "Blues")[2], span = 0.01) +
geom_line(aes(y=theta_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_1),], color=brewer.pal(5, "Blues")[1], alpha = 0.5) + geom_smooth(aes(y=theta_absolute_1, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_1),], color=brewer.pal(5, "Blues")[1], span = 0.01)

ggplot() + geom_line(aes(y=alpha_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_2),], color=brewer.pal(5, "Blues")[5], alpha = 0.5) + geom_smooth(aes(y=alpha_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_2),], color=brewer.pal(5, "Blues")[5], span = 0.01) +
geom_line(aes(y=beta_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_2),], color=brewer.pal(5, "Blues")[4], alpha = 0.5) + geom_smooth(aes(y=beta_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_2),], color=brewer.pal(5, "Blues")[4], span = 0.01) +
geom_line(aes(y=gamma_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_2),], color=brewer.pal(5, "Blues")[3], alpha = 0.5) + geom_smooth(aes(y=gamma_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_2),], color=brewer.pal(5, "Blues")[3], span = 0.01) +
geom_line(aes(y=delta_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_2),], color=brewer.pal(5, "Blues")[2], alpha = 0.5) + geom_smooth(aes(y=delta_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_2),], color=brewer.pal(5, "Blues")[2], span = 0.01) +
geom_line(aes(y=theta_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_2),], color=brewer.pal(5, "Blues")[1], alpha = 0.5) + geom_smooth(aes(y=theta_absolute_2, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_2),], color=brewer.pal(5, "Blues")[1], span = 0.01)

ggplot() + geom_line(aes(y=alpha_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_3),], color=brewer.pal(5, "Blues")[5], alpha = 0.5) + geom_smooth(aes(y=alpha_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_3),], color=brewer.pal(5, "Blues")[5], span = 0.01) +
geom_line(aes(y=beta_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_3),], color=brewer.pal(5, "Blues")[4], alpha = 0.5) + geom_smooth(aes(y=beta_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_3),], color=brewer.pal(5, "Blues")[4], span = 0.01) +
geom_line(aes(y=gamma_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_3),], color=brewer.pal(5, "Blues")[3], alpha = 0.5) + geom_smooth(aes(y=gamma_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_3),], color=brewer.pal(5, "Blues")[3], span = 0.01) +
geom_line(aes(y=delta_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_3),], color=brewer.pal(5, "Blues")[2], alpha = 0.5) + geom_smooth(aes(y=delta_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_3),], color=brewer.pal(5, "Blues")[2], span = 0.01) +
geom_line(aes(y=theta_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_3),], color=brewer.pal(5, "Blues")[1], alpha = 0.5) + geom_smooth(aes(y=theta_absolute_3, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_3),], color=brewer.pal(5, "Blues")[1], span = 0.01)

ggplot() + geom_line(aes(y=alpha_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_4),], color=brewer.pal(5, "Blues")[5], alpha = 0.5) + geom_smooth(aes(y=alpha_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$alpha_absolute_4),], color=brewer.pal(5, "Blues")[5], span = 0.01) +
geom_line(aes(y=beta_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_4),], color=brewer.pal(5, "Blues")[4], alpha = 0.5) + geom_smooth(aes(y=beta_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$beta_absolute_4),], color=brewer.pal(5, "Blues")[4], span = 0.01) +
geom_line(aes(y=gamma_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_4),], color=brewer.pal(5, "Blues")[3], alpha = 0.5) + geom_smooth(aes(y=gamma_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$gamma_absolute_4),], color=brewer.pal(5, "Blues")[3], span = 0.01) +
geom_line(aes(y=delta_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_4),], color=brewer.pal(5, "Blues")[2], alpha = 0.5) + geom_smooth(aes(y=delta_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$delta_absolute_4),], color=brewer.pal(5, "Blues")[2], span = 0.01) +
geom_line(aes(y=theta_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_4),], color=brewer.pal(5, "Blues")[1], alpha = 0.5) + geom_smooth(aes(y=theta_absolute_4, x = timestamps), 
       data = eeg[!is.na(eeg$theta_absolute_4),], color=brewer.pal(5, "Blues")[1], span = 0.01)

temp <- print(length(eeg[!is.na(eeg$blink), ]))

[1] 101